Retrospective Characterization of Clot Waveform Analysis in Canine Immune‐Mediated Hemolytic Anemia Using the Turbidimetric ACL‐TOP 300 Analyzer (2018–2022): 35 Cases

ABSTRACT

Objective

To characterize clot waveform analysis (CWA) parameters (delta and derivative values), alongside standard coagulation parameters (prothrombin time [PT], activated partial thromboplastin time [aPTT], and fibrinogen) in dogs diagnosed with immune‐mediated hemolytic anemia (IMHA) using a turbidimetric optical clot detection‐based analyzer.

Design

Retrospective study from January 2018 to October 2022.

Setting

University teaching hospital.

Animals

Thirty‐five client‐owned dogs diagnosed with IMHA.

Measurements and Main Results

CWA parameters were characterized by an increased PT delta in 35 of 35 dogs with IMHA (100%) and an increased PT first derivative, aPTT delta, and aPTT second derivative in 34 of 35 (97.1%). Compared with healthy controls, the PT was prolonged and the fibrinogen concentration was increased in most of the IMHA cohort, while the aPTT fell within the reference interval. Of the 18 patients that had D‐dimers performed, 17 (94.4%) measured above the reference interval.

Conclusions

CWA of patients with IMHA revealed increased parameters compared with the reference intervals, consistent with a hypercoagulable state. This measurement may be helpful in identifying patients that may benefit from antithrombotic therapy.

Abbreviations

aPTT

activated partial thromboplastin time

CWA

clot waveform analysis

IMHA

immune‐mediated hemolytic anemia

PT

prothrombin time

QFA

quantitative fibrinogen assay

RI

reference interval

ROTEM

rotational thromboelastometry

TE

thromboembolism

TEG

thromboelastography

VCM

viscoelastic coagulation monitor

1. Introduction

Coagulation disorders present significant risks and can cause life‐threatening complications in veterinary patients, necessitating early detection for effective prevention or treatment. Standard benchtop coagulation panels that perform prothrombin time (PT), activated partial thromboplastin time (aPTT), and quantitative fibrinogen assay (QFA) are widely available, inexpensive, and automated, and can help detect coagulopathies by identifying deficiencies in factor and fibrinogen levels. However, they do not provide information on the clotting process in vivo, and their value in predicting thromboembolic risk in patients with hypercoagulable states is limited.

Benchtop (thromboelastography [TEG], rotational thromboelastometry [ROTEM]) and point‐of‐care (TEG 6s¹, viscoelastic coagulation monitor [VCM]) viscoelastic coagulation testing equipment can rapidly measure and graphically display changes in whole blood viscoelasticity from the time a clot develops through its breakdown [1, 2, 3, 4]. This information can help guide the treatment of a patient's hemostatic deficiency appropriately by graphically revealing the contribution of clotting factors, platelets, and fibrinogen to overall clot strength. With the exception of the TEG 6s, all of these devices are validated in dogs [2, 3, 5]. However, these analyzers are prone to preanalytical error due to interference in cases of icterus, lipemia, or preactivation; require experienced personnel to operate; suffer from interdevice variability; and thus require institution‐specific reference intervals (RIs) for accurate interpretation [5, 6, 7, 8].

Clot waveform analysis (CWA) is an alternative method of coagulation testing based on optical detection of temporal changes in light transmittance and absorbance throughout the entire process of clot formation. The information provided includes the standard clotting times (PT/aPTT), fibrinogen, D‐dimers, a graphic display of clot curve/waveform, and calculations of associated values to interpret clot strength (delta), clot velocity (first derivative), and clot acceleration (second derivative) [9, 10]. In human medicine, CWA has been shown to be a reliable test in both hypocoagulable and hypercoagulable states, demonstrating a decrease in delta and derivative values in hypocoagulable states and an increase in hypercoagulable states [11, 12]. CWA has recently been validated for use on canine and feline samples, and while RIs for delta and derivative values were reported, institutional and device‐specific RIs are recommended [13]. A follow‐up study then characterized the delta and derivative values for common canine diseases that generally predispose to a hypocoagulable state [14]. To the authors’ knowledge, no studies in veterinary medicine have characterized the values for a hypercoagulable state.

Immune‐mediated hemolytic anemia (IMHA) has been documented as a disease that causes hypercoagulability [8, 15, 16, 17, 18, 19]. The mechanism by which this occurs is complex and multifactorial. It has been suggested that dysregulation of blood cells, coagulation factors, and cytokines contributes to the hypercoagulable state as well as secondary effects of hemolysis and changes in gene expression of both procoagulant and anticoagulant mediators [20, 21, 22, 23, 24, 25, 26, 27]. The clinical relevance of this hypercoagulable state is demonstrated in the development of thromboembolism (TE) in dogs with IMHA, which has been associated with increased mortality [15, 27, 28]. The ability to identify which patients may benefit from earlier administration of antithrombotic medications could potentially improve overall patient outcomes.

The objective of the current study was to characterize delta and derivative values in dogs diagnosed with IMHA and compare results with those from healthy controls. The hypothesis was that CWA parameters in canine patients with IMHA would be increased compared to normal RIs. The goal was to examine the PT, aPTT, and QFA values, as well as clot curve characteristics of high and low amplitude, delta, the first derivative of the PT curve, and the second derivative of the aPTT curve of dogs diagnosed with IMHA.

2. Materials and Methods

2.1. Animal Recruitment and Record Review

Medical records were searched from January 2018 to October 2022. Preliminary patient data were collected based on the diagnosis in the patient's electronic medical record (“anemia,” “anemia‐hemolytic,” “autoimmune hemolytic anemia,” and “immune mediated hemolytic anemia”). Dogs over 6 months of age were considered for inclusion. The preliminary data were further reviewed for specific inclusion criteria, which were the presence of diagnostic or supportive evidence of IMHA based on the 2019 American College of Veterinary Internal Medicine consensus statement and coagulation profile results reported at the time of diagnosis [29]. Only cases with coagulation curves that met the manufacturer's recommendations for adequacy were included. Exclusion criteria were treatment with immunosuppressive medication within 30 days before a coagulation panel was performed (regardless of indication) and previous blood transfusion within 6 months of a coagulation panel, regardless of indication. A healthy reference population of 48 previously described dogs served as the control group [13].

2.2. Analyzers

A turbidimetric optical clot detection‐based analyzer² was maintained and used per manufacturer's recommendations throughout the study period. Manufacturer reagents³ for PT, aPTT, QFA, and D‐dimers were used, and quality assurance was performed daily using high, low, and normal controls. The same reagents had been used throughout the study period.

All fibrinogen concentration values below the lower limit of linearity of 35 µmol/L (mg/dL) were set as 35 µmol/L (mg/dL). All fibrinogen concentration values above the upper limit of linearity of 1000 µmol/L (mg/dL) were set as 1000 µmol/L (mg/dL). One fibrinogen value failed automated QFA testing; instead, the fibrinogen was calculated using the reagent for PT per laboratory protocol. The upper limit of linearity for this calculated value was 700 µmol/L (mg/dL), and this data point was set at 700 µmol/L (mg/dL). The minimum and maximum absorbance values for the PT and aPTT curves were calculated by the analyzer, with the exception of two patients with values determined by visual inspection due to missing analyzer data. The delta values were then calculated by subtracting the minimum from the maximum. The first and second derivative values for the PT and aPTT, respectively, were recorded based on analyzer calculations.

Benchtop CBC⁴ and serum biochemical testing⁵ were submitted within 24 h of coagulation testing for all but one patient with IMHA and were performed per manufacturer's guidelines. PCV was obtained via point‐of‐care HCT centrifuge⁶. In the cases in which this was not performed, the HCT calculated from the CBC was used, if available.

2.3. Sample Collection

Samples were collected throughout the study period via institutional laboratory standard operating procedure into 1.8‐mL 3.2% sodium citrate vacutainer tubes⁷. Overfilled or underfilled tubes were rejected. Samples were centrifuged at 150 × g for 15 min and directly loaded onto the analyzer.

2.4. Statistical Analysis

For each variable of interest (PT, aPTT, QFA, D‐dimer, PT delta, PT first derivative, aPTT delta, aPTT second derivative), the proportion outside the previously reported 95% RI was calculated, with 95% CIs computed using the Clopper–Pearson method. Additionally, significance tests were performed against a null of 2.5% of the data being outside the interval using the exact binomial test. Wilcoxon tests were performed to compare the IMHA data with the previously collected control data, as some variables were not normally distributed or had outliers. All calculations were performed with freely available statistical software⁸, with a significance level set at 0.05.

3. Results

3.1. Animals

Thirty‐five dogs met the inclusion criteria, including 19 neutered males, 15 neutered females, and one intact male. The mean age was 7.5 years (range: 3.5–12 years). Represented breeds included five Labrador Retrievers, five mixed breed dogs, three Spaniel breeds, three Shih Tzus, two Yorkshire Terriers, two American Pit Bull Terriers, two German Shepherd Dogs, two Dachshunds, and one each of the following breeds: Pug, Pomeranian, Mastiff, Maltese, Greyhound, Golden Retriever, Doberman Pinscher, Pembroke Welsh Corgi, Boxer, Beagle, and Australian Shepherd.

3.2. Coagulation Data

Compared with control dogs, most dogs with IMHA had a prolonged PT and increased QFA, with 29 of 35 dogs (82.8%, 95% CI: 66.4%–93.4%) and 33 of 35 (94.3%, 95% CI: 80.8%–99.3%), respectively, falling above the respective RI (Table 1). The aPTT was above the RI in 11 of 35 dogs with IMHA (31.4%, 95% CI: 16.9%–49.3%) (Table 1), while 21 of 35 dogs (60%) fell within the RI.

TABLE 1.

Standard coagulation testing data for 48 healthy control dogs and 35 dogs with IMHA.

Variable	Group
	Healthy control, median	IMHA, median	RI
PT (s)	7.1	9.7	6.1–8.2
% above RI	—	82.9	—
aPTT (s)	15.5	17.7	12.4–18.7
% above RI	—	31.4	—
QFA (µmol/L, mg/dL)	154	762	98–232
% above RI	—	94.3	—
PLT (× 109/L, × 103/µL)	—	157	129–395
% below RI	—	37.1	—

Abbreviations: aPTT, activated partial thromboplastin time; IMHA, immune‐mediated hemolytic anemia; PLT, platelet count; PT, prothrombin time; QFA, fibrinogen concentration; RI, reference interval.

All CWA parameters (PT delta, PT first derivative, aPTT delta, aPTT second derivative) were generally outside of generated RIs in the dogs with IMHA compared with controls. For instance, 35 of 35 (100%, 95% CI: 90.0%–100.0%) of PT deltas and 34 of 35 (97.1%, 95% CI: 85.1%–99.9%) of each PT first derivatives, aPTT deltas, and aPTT second derivatives were above the RI (Table 2).

TABLE 2.

Clot curve data for 48 healthy control dogs and 35 dogs with IMHA.

Variable	Group
	Healthy control, median	IMHA, median	RI
PT high amplitude (mAbs)	128	309	45.7–528.7
PT low amplitude (mAbs)	96	32	13.1–465.6
PT delta (mAbs)	35	284	21.6–58.4
% above RI	—	100.0	—
PT first derivative (mAbs/s)	149.3	1003.408	76.6–271.7
% above RI	—	97.1	—
aPTT high amplitude (mAbs)	163	525	57.8–733.7
aPTT low amplitude (mAbs)	110	36	14.0–695.3
aPTT delta (mAbs)	53	501	30.6–94.5
% above RI	—	97.1	—
aPTT second derivative (mAbs/s2)	363.0	1981.777	190.3–600.0
% above RI	—	97.1	—

Abbreviations: aPTT, activated partial thromboplastin time; IMHA, immune‐mediated hemolytic anemia; PT, prothrombin time; RI, reference interval.

Figure 1 contains distributions for the standard coagulation panel data as well as the delta and derivative data from the clot waveforms. The differences in PT, aPTT, QFA, PT delta, PT first derivative, aPTT delta, and aPTT second derivative between healthy control dogs and dogs with IMHA were significant (Wilcoxon test; p < 0.05).

FIGURE 1.

Box plot distribution of standard coagulation tests, clot waveform analysis, platelet count, and PCV for 48 healthy control dogs and 35 dogs with IMHA. The line within each box represents the median. The lower and upper limits of each box plot represent the 25th and 75th percentile, respectively. The solid lines represent the reference intervals, and the dotted lines represent the confidence intervals around the reference bounds, as previously reported [13]. An asterisk (*) indicates statistically significant difference between IMHA and control dogs at p < 0.05, using the Wilcoxon test. The percentages above and below the RI limits are reported if significantly different from the expected 2.5% (with CIs). aPTT, activated partial thromboplastin time; CON, control; IMHA, immune‐mediated hemolytic anemia; PT, prothrombin time; QFA, fibrinogen concentration; first deriv, first derivative; second deriv, second derivative.

In addition to the standard coagulation panel (PT, aPTT, QFA), 18 of 35 patients also had D‐dimers measured. Of these patients, 17 of 18 (94.4%, 95% CI: 72.7%–99.9%) had D‐dimers measured above the RI. The difference in D‐dimers between dogs with IMHA and controls was statistically significant (Wilcoxon test; p < 0.05) (Figure 1).

4. Discussion

This retrospective study is the first to report CWA parameters derived from a turbidimetric optical coagulation analyzer to characterize clot strength (delta), clot velocity (first derivative), and clot acceleration (second derivative) in dogs with IMHA.

In our study, CWA found that 35 of 35 patients with IMHA (100%) had an increase in PT delta and 34 of 35 (97.1%) an increase in aPTT delta, indicating that these patients formed more robust clots with increased density. In addition, 97.1% of patients with IMHA had increased first and second derivatives, indicating an increased velocity and acceleration of clot formation. These data support our hypothesis that patients with IMHA have increased CWA parameters compared with the RI.

Several studies in people have investigated CWA parameters in patients with suspected or known hypercoagulable disease states (venous TE, acute cerebral infarction, COVID‐19) [12, 30, 31]. In each of these studies, delta or derivative values were increased when compared with healthy controls, which demonstrates rapid formation of strong clots, as expected in a hypercoagulable state [11, 12, 32]. Similar to findings in the human literature, the increases in delta and derivative values in our IMHA cohort support the presence of a hypercoagulable state.

Increased fibrinogen and D‐dimer values are frequently seen in patients with hypercoagulable states. In our population of patients with IMHA, as in other canine studies of IMHA, increased fibrinogen and D‐dimer concentrations were also found [17, 18, 23, 25, 33]. Studies report this is also commonly found in human patients with hypercoagulable states [11, 30, 32]. It is well established that coagulation and inflammation are closely associated; previous studies have demonstrated that fibrinogen, which is positively correlated with the acute‐phase C‐reactive protein, may be a marker of inflammation [18, 23, 34]. D‐dimers are a product of fibrin degradation and, when increased, indicate clot formation and subsequent breakdown. Therefore, the increased fibrinogen and D‐dimer concentrations in this study population were expected.

The majority of patients in the current study had a prolonged PT (29/35 [82.9%]) and a normal aPTT (21/35 [60%]). Prolonged PT or aPTT values suggest deficiencies in specific clotting factors either by decreased production, rapid consumption, or factor dysfunction, leading to a possible hypocoagulable state; however, CWA parameters from these patients suggest they are, in fact, hypercoagulable [35, 36]. There are mixed results in human and veterinary studies regarding the use of PT and aPTT as reliable predictors of hypercoagulability and the risk of thrombus formation; when comparing hypercoagulable TEG tracings to standard coagulation tests (PT/aPTT), studies have shown PT or aPTT to be shortened, normal, or prolonged [35, 36, 37, 38, 39, 40]. In addition, patients with normocoagulable TEG tracings can have alterations in PT or aPTT [41, 42]. Other studies have shown abnormally short PT or aPTT values to be associated with increased risk of TE [37, 43, 44]. Our data provide further evidence that standard coagulation tests cannot reliably predict hypercoagulability.

Apart from standard benchtop coagulation panels, the other commonly used analyzers for assessing coagulation status in veterinary medicine include TEG, ROTEM, and VCM. Each of these modalities graphically displays changes in blood viscoelasticity from clot development to breakdown, thereby providing additional information on the underlying mechanisms of coagulopathy in individual patients [1, 2, 4]. TEG and ROTEM have been used in human and veterinary medicine for many years; therefore, more published data are available. However, TEG and ROTEM require experienced personnel to use and are prone to preanalytical error [2, 6]. VCM is becoming more widely available and is easy to use; however, as a newer technology, there are currently only a few publications in veterinary medicine evaluating its effectiveness in detecting coagulopathies. One study found VCM to be sensitive in diagnosing normal coagulability, but its specificity was moderate to poor, misclassifying hypocoagulable and hypercoagulable dogs with normal coagulability in up to 44% of cases when compared with ROTEM [7]. In another VCM study, results found up to 24% of healthy controls were falsely classified as hypocoagulable when compared with TEG [8]. Additionally, interdevice correlation is variable, and individual RIs are recommended for interpreting results across devices [5, 7, 8]. Turbidimetric coagulation analyzers and CWA remain relatively novel techniques in veterinary medicine, although they are widely used in human medicine for detecting both hypercoagulable and hypocoagulable states [11, 12, 45, 46]. CWA provides the standard clotting values and clot dynamics presented as measurable parameters and graphic displays, although graphic displays were not specifically used in this study. Because plasma, rather than whole blood, is used, preanalytical factors such as icterus, lipemia, and preactivation have not been shown to significantly affect results [13]. Further studies are warranted to evaluate the graphic displays produced by this technology, which could enhance our ability to interpret results and direct patient therapy.

IMHA was chosen as the representative disease in this study based on numerous previous studies demonstrating a hypercoagulable state in affected dogs [15, 16, 17, 18]. Hypercoagulability in patients with IMHA is likely multifactorial, with suggested mechanisms including increased activity of monocytes, neutrophils, leukocytes, and proinflammatory cytokines and an increase in circulating activated platelets [20, 21, 23, 24, 25, 27]. Tissue factor is a major contributor to coagulation, and its activity and expression are upregulated in patients with IMHA [25, 47]. Hemolysis results in the release of free hemoglobin and microparticles, both of which further enhance inflammation and activation of coagulation [47, 48, 49, 50]. Upregulation in gene expression of procoagulant mediators, as well as downregulation in gene expression of anticoagulant mediators, has also been demonstrated in patients with IMHA [25, 26]. Although venous TE appears to be more common, arterial TE has been documented in patients with IMHA, further suggesting that the hypercoagulable state of IMHA is multifactorial and includes dysregulation of platelets as well [27, 28, 51, 52]. These findings explain the current recommendation to use both anticoagulants and antiplatelet medications to prevent thrombotic complications in dogs with IMHA that have an adequate platelet count and no evidence of hypocoagulability [51].

Most dogs included in this study were neutered (34/35 [97.1%]), which is consistent with previous studies describing the incidence of IMHA being higher in neutered dogs [22, 52]. However, many studies describe a higher incidence of IMHA in female dogs, which was not true for our study population, in which males predominated (20/35 [57.1%]) [5, 8, 53, 54]. In contrast, other studies have reported populations with similar numbers of males and females [16, 27]. Although IMHA can occur in any breed, previously documented overrepresented breeds include Cocker and Springer Spaniels, Poodles, Sheepdogs, Irish Setters, and Collies [15, 16, 17, 18, 27, 28, 52, 55]. The most prevalent breeds in our study were Labrador Retrievers (5/35 [14.3%]) and mixed‐breed dogs (4/35 [11.4%]). Only three dogs (8.6%) were of the Spaniel variety, including two English Springer Spaniels and one Cocker Spaniel. However, we did not compare breed prevalence in this study with hospital breed representation over the same period, so no conclusions should be drawn regarding IMHA breed predisposition from these data. The median age of included patients was 7.5 years, which is consistent with previous studies in canine IMHA [15, 16, 17, 18, 27, 28, 52, 55].

There are several limitations to this study. Patient records may be incomplete in a retrospective study, which may have introduced case identification or selection bias. The patient population was from a single institution and was also small because the specified analyzer had only been in use since 2018. A source of bias inherently comes with the diagnosis of IMHA. Nonassociative (formerly primary) IMHA is a diagnosis of exclusion, whereas associative (formerly secondary) IMHA is associated with an underlying condition such as infectious disease or neoplasia. Many of the underling conditions of associative IMHA may themselves induce a proinflammatory state, which has been associated with hypercoagulability [29]. Also, dogs with IMHA that also had potentially prothrombotic comorbidities such as hyperadrenocorticism, pancreatitis, and neoplasia were not excluded. Despite the current CURATIVE guidelines stating that thrombosis secondary to these conditions is associated with only a small subset of dogs, we cannot determine whether hypercoagulable CWA parameters demonstrated in these dogs was related to IMHA or comorbidity [56].

Additional limitations exist in the use of the American College of Veterinary Internal Medicine consensus statement and its associated algorithm for the diagnosis of IMHA, as bias cannot be excluded and there is room for interpretation since it is intended for clinical rather than research purposes. For example, a sign of immune‐mediated RBC destruction is the presence of spherocytes, though there are also nonimmunologic primary causes of spherocytosis, such as oxidative damage from zinc intoxication, envenomation, and splenic torsion. Identifying whether spherocytosis is nonimmunologic or due to IMHA cannot be distinguished cytologically [57, 58, 59, 60]. Information from the patient records may have been missed or misinterpreted when going through the consensus algorithm, which could have further led to the false inclusion or exclusion of patients. The presence of anemia has been associated with hypercoagulability. The suggested mechanisms for in vitro hypercoagulable test results associated with anemia include an increase in the relative amount of plasma proteins to RBCs, a decreased citrate‐to‐plasma protein ratio, and a decreased RBC mass allowing for a tighter fibrin meshwork [61, 62]. Due to the inclusion criteria for this study, all patients were anemic. Therefore, anemia may have contributed to the altered CWA parameters and the depiction of hypercoagulability in our study, as we did not compare patients with IMHA to those with other causes of anemia. Finally, clinical evidence of TE was not accounted for in this patient population, and no necropsies were evaluated to assess for signs of TE. Despite the CWA parameters supporting a hypercoagulable state in this population, we are unable to comment on any association between our findings and the likelihood of TE.

The current study is the first to characterize CWA parameters in dogs with IMHA, a disease documented to cause a hypercoagulable state. The results of this study suggest that CWA parameters in dogs with IMHA are consistent with hypercoagulability and that they may be used to identify dogs with greater thromboembolic risk. Additional studies are needed to confirm these findings and to attempt to identify a relationship between such CWA findings and the development of thrombosis.

Author Contributions

Leah G. Richter: writing – original draft, writing – review and editing, conceptualization, investigation, data curation. Tasia M. Ludwik: conceptualization, investigation, writing – review and editing, supervision. Angela D. Gwynn: writing – review and editing, methodology, conceptualization, data curation. Aaron Rendahl: investigation, formal analysis, methodology, software, writing – review and editing. Jeffrey M. Todd: conceptualization, investigation, writing – review and editing, supervision.

Ethics Statement

The authors confirm that the ethical policies of the Journal, as noted on the Journal's author guidelines page, have been adhered to and the appropriate ethical review committee approval has been received. The US National Research Council's guidelines for the Care and Use of Laboratory Animals were followed.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.